Great submission. This claim, though... "Traditionally, achieving all of these goals has required an expensive combination of atomic commitment and replication

Great submission. This claim, though... "Traditionally, achieving all of these goals has required an expensive combination of atomic commitment and replication protocols – introducing extensive coordination overhead. Our system, Eris, takes a different approach. It moves a core piece of concurrency control functionality, which we term multi-sequencing, into the datacenter network itself. This network primitive takes on the responsibility for consistently ordering transactions, and a new lightweight transaction protocol ensures atomicity." ...reminds me of NASA's ROBUS and SPIDER: https://shemesh.larc.nasa.gov/fm/spider/ https://shemesh.larc.nasa.gov/fm/spider/ So, moving integrity and availability mechanisms to hardware in the middle is a proven method. They even did a proof of it. Good to see that kind of thinking is being applied to networking hardware. A tech like this might have most impact at companies like Google that own some of the lines they use. Might even be a value-added service for Tier 1's if we're lucky.